Overview of the prediction framework. Diagram representing the workflow of bioinformatic prediction of putative OMPs from the avian and porcine strain genomes of P. multocida. Ten predictors were categorized into 3 groups: subcellular localization, transmembrane β-barrel protein prediction and outer membrane lipoprotein prediction. The predicted proteins in each group were filtered by consensus prediction and combined to form a single integrated list. Text mining and sequence analyses were used to confirm that the predicted proteins were outer membrane-associated with a high degree of confidence. The proteins filtered-out from the avian strain genome were also analyzed further by data mining and sequence analysis. The numbers of predicted proteins in each step are shown in parentheses: the first number represents proteins from the avian strain genome and the second number from the porcine strain genome.